Design And Implementation Of City Commercial Bank Credit System Under Internet Finance

The business model and processing flow of Internet credit have brought certain challenges to the original business system.The traditional service-oriented SOA architecture,RPC or MVC architecture system is difficult to support the existing business scenarios with multiple scenarios and frequent product changes.Taking the most important credit business of the bank as an example,its characteristics include large business concurrency,fast approval process,less manual intervention,and more post-loan management is done automatically by the system.If we only rely on the transformation of the original personal loan management system to achieve the Internet "inclusive finance" model,it will be difficult to achieve significant development.In addition,since the business is mainly run on the Internet,the prevention and control of credit risk is even more important.It is difficult to prevent various risks from the Internet by continuing to adopt the traditional expert experience approval mode.Based on the separation of microservices and the management of microservices in the bank credit system,this article introduces container virtualization technology,combines Docker container technology and Kubernetes container management components,and explores for convenient development,clear function module boundaries,and low coupling between modules.System architecture that facilitates horizontal expansion and deployment.Introduced Drools decision engine technology and improved GBDT integrated learning algorithm for business approval efficiency and risk prevention and control,which solves the problems faced by the Internet credit business as a whole.The implementation of the system will play an important role in the future development of the city's credit business Practical significance.The main work of this article includes the following aspects:(1)Based on the design concept of microservice,the traditional single system is divided,and the divided service is described in detail.In terms of non functional requirements,the design standards of reliability and safety are proposed.This paper analyzes the main problems of Internet credit business,gives corresponding solutions to the problems,and analyzes and studies the technology and calculation used in the solutions.Including the Rete algorithm and operation principle of decision engine technology,as well as the research and practice of improved GBDT integrated learning algorithm used in personal credit risk assessment.(2)The design and implementation of the system.The design mainly includes the general principles of the design,the design of the Dubbo-based application architecture,the functional design,the detailed system design,and the database design.The application architecture is mainly based on the Dubbo standard,designing the microservice architecture of the Internet credit system and the technical middleware required for microservice governance;the system architecture design is based on Kubernetes and Docker technologies,and makes preliminary plans for the containerization and deployment of the application;detailed design Based on the overall design,the system's main functional modules,business models,and data flow are designed.In the system implementation part,the establishment of Kubernetes environment and the deployment of containers are introduced.In addition,a detailed description is given for the system integrated with the microservice application involved in the functional requirement analysis.Includes the main page view of each functional module,the business data to be displayed in the module,the user can add,delete,modify,and check the operations for these data,and the corresponding changes in the data after the operation,etc.(3)The system test mainly describes the minimum software and hardware requirements to ensure the normal operation of the system,the purpose of the test,the functional scope of the test,and the goals to be achieved.